postlasso

Runs Lasso over the candidate predictors, keeps predictors with nonzero selected coefficients,

and refits an ordinary least-squares model on the selected predictors for coefficient inference.

If no predictors are selected, postlasso fits an intercept-only model unless noconstant is

specified.

Syntax

postlasso linear y x1 x2 [, alpha(<num>) robust noconstant]

Examples

postlasso linear wage educ exper tenure
postlasso linear wage educ exper tenure, alpha(0.05)
postlasso linear wage educ exper tenure, robust
postlasso linear wage educ exper tenure, noconstant

See also

• Command Reference Index
• User Guide
• In-App Help System